MURDOCH RESEARCH REPOSITORY
This is the author’s final version of the work, as accepted for publication following peer review but without the publisher’s layout or pagination.
The definitive version is available at http://dx.doi.org/10.1016/j.micromeso.2012.10.027
Huang, W-Y, Li, D., Yang, J., Liu, Z-Q, Zhu, Y., Tao, Q., Xu, Kai, Li, J-Q and Zhang, Y-M (2013) One-pot synthesis of Fe(III)-
coordinated diamino-functionalized mesoporous silica: Effect of functionalization degrees on structures and phosphate
adsorption. Microporous and Mesoporous Materials, 170 . pp. 200-210.
http://researchrepository.murdoch.edu.au/13060/
Copyright: © 2012 Elsevier Inc
It is posted here for your personal use. No further distribution is permitted.
Accepted Manuscript
One-pot synthesis of Fe(Ш)-coordinated diamino-functionalized mesoporous
silica: Effect of functionalization degrees on structures and phosphate adsorp‐
tion
Wei-Ya Huang, Dan Li, Jun Yang, Zhao-Qing Liu, Yi Zhu, Qi Tao, Kai Xu,
Jian-Qiang Li, Yuan-Ming Zhang
PII: S1387-1811(12)00688-9
DOI: http://dx.doi.org/10.1016/j.micromeso.2012.10.027
Reference: MICMAT 5822
To appear in: Microporous and Mesoporous Materials
Received Date: 3 July 2012
Revised Date: 12 August 2012
Accepted Date: 17 October 2012
Please cite this article as: W-Y. Huang, D. Li, J. Yang, Z-Q. Liu, Y. Zhu, Q. Tao, K. Xu, J-Q. Li, Y-M. Zhang, One-
pot synthesis of Fe(Ш)-coordinated diamino-functionalized mesoporous silica: Effect of functionalization degrees
on structures and phosphate adsorption, Microporous and Mesoporous Materials (2012), doi: http://dx.doi.org/
10.1016/j.micromeso.2012.10.027
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
One-pot synthesis of Fe(Ш)-coordinated
diamino-functionalized mesoporous silica: Effect of
functionalization degrees on structures and phosphate
adsorption
Wei-Ya Huang a,b, Dan Lic, Jun Yang a *, Zhao-Qing Liua, Yi Zhua, Qi Taod, Kai Xu a, Jian-Qiang
Li a, Yuan-Ming Zhang a *
( aDepartment of Chemistry, Jinan University, Guangzhou, 510632, China; bDepartment of
Materials Science and Engineering, Taizhou University, Linhai, 317000, China; cEnvironmental
Engineering, School of Environmental Science, Murdoch University, Murdoch, Western Australia,
6150, Australia; dKey Laboratory of Mineralogy and Metallogeny, Guangzhou Institute of
Geochemistry, Chinese Academy of Sciences, Guangzhou, 510460, China )
* To whom correspondence should be addressed. E-mail: [email protected] (Y.M. Zhang);
[email protected] (J. Yang); Fax: +86-20-85220014; Tel: +86-20-85221264.
2
Abstract:
Fe(Ш)-coordinated mesoporous silica adsorbents functionalized with different loadings of
diamino groups were prepared by a new NH4F-assisted co-condensation method and impregnation
of Fe3+ cations. Various characterization techniques, e.g. XRD, SEM, TEM, ICP-MS, elemental
analysis, FT-IR spectroscopy, and nitrogen adsorption-desorption, were utilized to investigate the
effect of functionalization degrees of absorbents on their chemical composition, surface chemistry,
pore structures and phosphate adsorption capacities in detail. In the batch adsorption tests, the
functionalized absorbents with increasing loadings of diamino groups possessed markedly
enhanced adsorption capacities, although there was a gradual loss of ordered mesostructures
accompanied. The adsorption isotherms were represented better by using Langmuir model than
Freundlich model, which indicated the presence of monolayer adsorption. In particular, for the
resulting absorbent prepared with 0.5:1 molar ratio of AAPTS and TEOS, the maximum
phosphate capture capacity calculated from Langmuir model is 20.7 mg P/g. In the kinetic study,
the phosphate adsorption followed pseudo-second-order equation well with a correlation
coefficient of 0.999, suggesting the adsorption process be chemisorption. The phosphate
adsorption efficiency of prepared absorbent was highly pH-dependent and the high removal of
phosphate was achieved within the pH between 3.0 and 6.0. The presence of Cl− and NO3−
exhibited small impacts on the phosphate adsorption by using our synthesized absorbent; whereas,
there were significantly negative effects from HCO3− and SO4
2− on the phosphate removal. In
0.010 M NaOH, more than 90% of the absorbed phosphate anions on the spent adsorbent could be
desorbed, suggesting the absorbent with a capacity of regeneration.
Keywords: diamino, SBA-15, functionalization, phosphate, adsorption
3
1. Introduction
Excessive presence of phosphate in water bodies can lead to a significant eutrophication
problem of rivers, lakes and sea. The negative environmental impacts of eutrophication include
overgrowth of algae and depletion of dissolved oxygen, which subsequently result in depopulation
of aquatic animals and acceleration of water scarcity [1]. Therefore, phosphate removal from
wastewater containing high concentrations of nutrients prior to its discharge is essential to
conserve aquatic environment. To date, there has been intensive effort devoted towards different
approaches, including biological treatment, adsorption, and chemical precipitation, to remove
phosphate from water [2]. In particular, adsorption method has been most widely studied in
phosphate removal, due to the high removal efficiency and fast removal rate, as compared with
those by utilizing chemical precipitation and biological treatment [2–4]. Hence, a variety of
adsorbents have been developed and evaluated for phosphate removal, including goethite [5],
palygorskite [6], vesuvianite [7], Fe oxide tailing [8], layered double hydroxides [9], calcite [10],
zeolite [4,11], pillared montmorillonite [12], red mud [13], fly ash [14], blast furnace slag [15],
collagen fiber [16], orange waste [17], etc. Until now, the desire to develop a novel adsorbent
which offers high effectiveness and low cost never stops.
Recently, the use of mesoporous silica materials as absorbents has attracted great interest,
due to their large surface areas, as well as controllable pore sizes and arrangements [18]. Organic
functionalization of mesoporous silica materials via covalently grafting organic functional groups
onto silica substrates can enhance their removal selectivity and efficiency to different
contaminants, even at a relatively low concentration [19–21]. In order to enhance the
accessibilities of molecules and absorption capacities for heavy metal cations [22, 23],
4
mesoporous silica materials with high loadings of functional groups and well-defined
mesochannels are highly desirable [24–27]. Different organic functional groups, i.e. amine, thiol,
carboxylic, and aromatic, etc., have been covalently bonded to the structures of mesoporous
materials via a post-synthesis grafting or co-condensation method. In particular, there has been
tremendous work on the development of amino-functionalized mesoporous materials for the
adsorptive removal of phosphate pollutant anions from water. For instance, Chouyyok and
co-workers reported the synthesis of Cu(II) and Fe(III)-coordinated ethylenediamine
(en)-functionalized MCM-41 mesoporous materials by using the post-synthesis grafting method.
Fe(III)-coordinated en-modified MCM-41 mesoporous silica was more effective to remove
phosphate anions, with the estimated maximum capacity 43.3 mg/g; whilst there was a rapid
sorption rate to remove 99% of phosphate within 1 min and the phosphate content in the solution
could be lowered to around 10 μg/L [28]. Zhang et al. investigated the adsorption behaviors of
phosphate with the use of Fe(III) and La(III)-coordinated diamino-functionalized MCM-41
adsorbents, which were fabricated by the post-synthesis grafting method. Their maximum
adsorption capacities could reach up to 51.8 mg/g and 54.3 mg/g, respectively [29, 30]. Long et al
utilized the post-synthesis grafting method to functionalize MCM-41, MCM-48 and SBA-15
materials with diamino groups. The adsorption capacity of Fe(III)-coordinated
diamino-functionalized MCM-41 (Fe-2N-MCM-41) was superior to those with other chelating
metal ions, including Cu2+, La3+, and Al3+. Furthermore, the adsorption capacities were observed
to increase in the order: Fe-2N-SBA-15 > Fe-2N-MCM-48 > Fe-2N-MCM-41[31]. Recently, a
series of Fe(III)-coordinated MCM-41 adsorbents with different organic-functionalization degrees
were prepared by the direct co-condensation method. The absorbent with a molar ratio of organic
5
functional groups to tetraethyl orthoslicate (TEOS) at 30% exhibited the maximum adsorption
capacity which was 52.5 mg/g [32].
To the best of our knowledge, until now, most of work focuses on the development of
organic-functionalized MCM-type mesoporous materials for phosphate removal; there were very
few investigations on the use of metal-coordinated diamino-functionalized SBA-15 mesoporous
silica. SBA-15, which is a member in the family of mesoporous molecular sieves, shows attractive
features, including hexagonally ordered mesopores and thick pore walls, as well as good thermal
and hydrothermal stability [33]. Moreover, in the above-mentioned literature, the post-synthesis
grafting method was commonly utilized to synthesize amino-functionalized MCM-type
mesoporous materials for phosphate removal. One of the drawbacks associated with this method is
an uncontrollable and heterogeneous distribution of organic moieties on the material surface. In
particular, organic moieties tend to congest near the entries of mesopores and reduce effective pore
sizes, resulting in a limited phosphate adsorption [34]. In contrast, the direct co-condensation can
provide a more homogeneous distribution of organic moieties in the framepores and the formation
of more accessible pores [34, 35]. Furthermore, the co-condensation method can potentially
minimize the number of processing steps and cost, thereby it is preferred over the post-synthesis
grafting method.
In this paper, we reported the synthesis of Fe(Ш)-coordinated amino-functionalized SBA-15
mesoporous materials via the direct co-condensation method and metal cation incorporation
process. It is known that the presence of organosilane in the synthesis gel would cause the
formation of partially or even completely disordered porous structures in the co-condensation
process [36–38], which affects the absorption capacities of resulting materials. In order to
6
overcome such a problem, fluoride ions were added to aid the synthesis of ordered mesoporous
adsorbents with a high density of functional groups via the direct co-condensation [39, 40]. The
structure properties of synthesized Fe(Ш)-coordinated absorbents with different
organic-functionalization degrees and their corresponding phosphate adsorption behaviors were
characterized in detail. In particular, the sample S15-NN-Fe-0.5, which was synthesized from the
starting gel with a molar ratio AAPTS/TEOS of 0.50, was studied on its phosphate capture in
aqueous solution by varying pH, ionic strength, and coexisting anions; as well as its desorption
kinetic in 0.010 mol/L NaOH solution.
2. Methods and materials
2.1 Synthesis of materials
Diamino-functionalized SBA-15 mesoporous silica materials were synthesized by using the
direct co-condensation method with the addition of ethylenediamine (en)-terminated organosilane
N-(2-aminoethyl)-3-aminopropyltrimethoxysilane (AAPTS, 95%, Aladdin Reagent Inc.). The
molar composition of reagents in the synthesis gel was a P123: b HCl: c H2O: d AAPTS: e TEOS
= 0.017: 6.3: 121.4: x: 1; in which x = 0, 0.10, 0.20, 0.30, 0.40, 0.50, and 0.60. Typically, 4.0 g of
EO20PO20EO20 (P123, Aldrich) was dissolved in the mixture with 30.0 g of deionized water and
120.0 g of 2.0 mol/L HCl solution under stirring at 35 °C for 3 h. After that, AAPTS and TEOS
(95%, Aladdin Reagent Inc.) were added into the mixture and stirred for 1 h, followed by the
addition of 133.0 mg NH4F (analytical grade, Guangzhou Chemical Reagent Factory). The
resulting solution was stirred at 35 °C for 20 h and subsequently transferred into a PTFE-lined
autoclave, which was heated at 100 °C for 24 h. The resulting solids were filtered, washed, and
7
dried at 60 °C. To remove surfactant P123, 1.0 g of the produced particles were suspended in a
solvent mixture consisting of 20.0 mL of ethanolamine and 80.0 mL of ethanol, which was
allowed to reflux for 24 h as reported [38]. After that, the solids were recovered by filtration,
washed with ethanol and stirred in 0.10 mol/L FeCl3 aqueous solution for 2 h. The final product
was separated from the filtrate, washed with deionized water and 2-propanol, and dried under
vacuum at 60 °C for 12 h.
The synthesized Fe(Ш)-coordinated diamino-functionalized powders were denoted as
S15-NN-Fe-x, where x refers to the molar proportion of AAPTS and TEOS in the synthesis gel,
ranging from 0 to 0.60. When the molar ratios of AAPTS/TEOS are 0, 0.10, 0.20, 0.30, 0.40, 0.50
and 0.60 in the gel during synthesis, the resulting samples are referred to S15-NN-Fe-0,
S15-NN-Fe-0.1, S15-NN-Fe-0.2, S15-NN-Fe-0.3, S15-NN-Fe-0.4, S15-NN-Fe-0.5, and
S15-NN-Fe-0.6, respectively.
2.2 Characterization of materials
Surface morphologies of samples were examined by scanning electron microscopy (SEM,
JSM-7401F, JEOL Ltd., Japan). X-ray powder diffraction (XRD) patters were recorded in the 2θ
range of 0.6–6 ° with a scan speed of 1 °/min by using a diffractometer (Bruker D8 Advance
diffractometer, Germany) with Cu Kα radiation (40 mA, 45 kV). Nitrogen adsorption-desorption
isotherms were measured at 77 K using ASAP 2010 (Micromeritics Inc., USA). Prior to analysis,
the samples were degassed at 120 °C for 12 h under vacuum. The specific surface area, SBET, was
determined from the linear part of the BET plot (P/P0 = 0.05-0.20). The pore size was calculated
from the desorption branch of isotherm by using Barrett-Joyner-Hallenda (BJH). The total pore
volume, Vtotal, was evaluated from the adsorbed nitrogen amount at a relative pressure 0.98.
8
Fourier transform infrared (FT-IR) measurements were performed between 400 cm−1 and 4000
cm−1 by using Shimadzu IR Prestige-21 instrument, in which KBr pellets containing 0.50% of the
samples were used. The N contents in samples were determined by elemental analysis with the use
of Series II CHNS/O Analyzer 2400. The degrees of N immobilization were calculated by
dividing the measured N contents in the elementary analysis by the theoretical N contents. The Fe
contents of samples were analyzed by Perkin-Elmer ICP-MS (model ELAN-DRC-e, USA). 0.10 g
of adsorbents were dispersed in 100.0 mL of 2.0 vol% HNO3 aqueous solution for 12 h and the
filtrates were used to determine the Fe contents in samples.
2.3 Phosphate adsorption experiments
A series of batch tests were conducted to investigate the phosphate adsorption performances
of absorbents. 0.10 g of absorbent was added into 100.0 mL of 0.065 mmol/L phosphate solution
in a polypropylene bottle, which was prepared by dissolving anhydrous K2HPO4 (analytical grade,
Sinopharm Chemical Reagent Co., Ltd) in deionized water. After shaken for 2 h at 35 °C, the
solution was removed by filtering through a syringe nylon-membrane filter (pore size 0.45 μm;
Shanghai Minglie Science Technology Co,. Ltd). The concentraiton of phosphate in filtrate was
analyzed by Autoanalyzer 3 (Bran and Luebbe Inc., Germany). The degree of removal was
calculated by Eq. (1) :
Degree of removal (%) = 1000
0 ×−C
CC e (1)
where C0 and Ce are the phosphate concentrations (mol/L) in the initial solution and filtrate,
respectively.
In equilibrium experiments, 0.050 g of absorbent was added into 100.0 mL of phosphate
solution prepared with various initial concentrations. The sealed polypropylene bottles were then
9
shaken at 35 °C for 2 h. The amount of phosphate adsorbed on the sample at the equilibrium (qe)
was calculated by,
m
VCCq e
e
×−= )( 0 (2)
where C0 and Ce are the initial and equilibrium phosphate concentrations in solution (mg P/L),
respectively; V is the volume of solution (L) and m is the mass of adsorbent (g).
The equilibrium data were fitted to the well-known Langmuir and Freundlich isotherms
models, as shown in Eqs. (3) and (4), respectively [41],
Langmuir model: 00
1q
C
Kqq
C e
Le
e += (3)
Freundlich model: eFe C
nKq log
1loglog += (4)
where Ce is the concentration of phosphate solution at equilibrium (mg P/L); qe is the
corresponding adsorption capacity (mg P/g); q0 (mg/g) and KL (L/mg) are constants in Langmuir
isotherm model which are related to adsorption capacity and energy or net enthalpy of adsorption,
respectively; KF (mg/g) and n are the constants in Freundlich isotherm model, which measure the
adsorption capacity and intensity, respectively.
Adsorption kinetic experiments were conducted as follows: 0.10 g of adsorbent was added in
200.0 mL of phosphate solution with different initial concentrations, e.g. 0.65 mmol/L, 1.63
mmol/L, or 2.60 mmol/L. The sealed polypropylene bottle was then placed in the shaker bath at
35 °C for 4 h. 2.0 mL of suspension was taken out of bottle over a given period of time to analyze
the phosphate concentration.
In order to analyze the kinetic mechanism of adsorption process, the experimental kinetic
data were fitted in the pseudo-second-order model, which are described as Eqs. (5) [42, 43]:
Pseudo second-order equation: eet q
t
qkq
t += 22
1 (5)
10
where qt and qe are the amount of phosphate adsorbed over a given period of time t (mg P/g) and
at equilibrium (mg P/g), respectively; and t is the sorption time (min); k2 is the equilibrium rate
constant of pseudo-second-order adsorption (g/mg/min).
To investigate the effect of pH on phosphate adsorption, 0.10 g of adsorbent was added in
100.0 mL of 0.065 mmol/L phosphate solution with different initial pH values, ranging from 2.0
to 12.0, at 35 °C. The initial pH of phosphate solution was adjusted with 1.0 mol/L NaOH and
HCl solution. The effect of coexisting anions on the degrees of phosphate removal was also
evaluated by dissolving sodium salt forms of F−, Cl−, NO3−, SO4
2− and HCO3− into 100.0 mL of
0.065 mmol/L phosphate solution, in which 0.10 g of adsorbent was added.
Desorption kinetic study was carried out to investigate the capacity of spent adsorbent’s
regeneration. First of all, 0.10 g of absorbent was placed in 100 mL of 6.45 mmol/L phosphate
solution at 35 °C for 2 h. After the adsorption−desorption equilibrium reached, the spent
absorbent was filter and washed carefully with deionized water to remove any unabsorbed
phosphate. Then the spent adsorbent was mixed with 100.0 mL of 0.010 mol/L NaOH solution
and the sample was taken out over a given period of time t. The phosphate desorption ratio was
estimated by,
Desorption ratio (%) = 100 ×
××
mq
VC
e
t (6)
where Ct is the phosphate concentration in filtrate over a given period of time (mg P/mL); qe is the
adsorbed phosphate amount on the spent adsorbent at the adsorption−desorption equilibrium (mg
P/g); V is the volume of NaOH solution (L) and m is the mass of spent adsorbent (g), respectively.
11
3. Results and discussion
Scheme 1 illustrates the possible mechanism on the preparation of Fe-coordinated
diamino-functionalized silica absorbents, as well as their phosphate absorption and desorption
processes. In the presence of AAPTS as organosilane in the synthesis gel, diamino-functionalized
silica absorbents can be fabricated via a new F--assisted co-condensation method. After the
removal of template (surfactant P123), the dimino-functional groups coordinate with Fe(Ш) to
form Fe3+–en complexes as trapping centers, onto which phosphate anions can directly bound [28].
The absorbed phosphate anions can be released in alkaline solution and thus the Fe-coordinated
diamino-functionalized silica absorbents show potential for regeneration and reuse.
3.1 Characterization of materials
The X-ray diffraction patterns of S15-NN-Fe-x (x = 0-0.6) samples, which are prepared with
different concentrations of AAPTS in the initial gel, are presented in Figure 1. Figure 1a shows the
XRD pattern of S15-NN-Fe-0, which is synthesized in a gel without adding AAPTS organosilane.
There are three well-resolved SBA-15 reflections at 2.. values of 0.5 ° --- 3 °, including one intense
diffraction peak (100) and two weak diffraction peaks (110) and (200), corresponding to the
presence of well-ordered hexagonal mesoporous silica framework. From Figure 1b to Figure 1f,
for the samples synthesized with AAPTS/TEOS molar ratios ranging from 0.10 to 0.50, the peak
intensity of (100) plane decreases gradually; and the (110) and (200) diffractions become less
resolved. In the XRD pattern of S15-NN-Fe-0.6 (Figure 1g), all of the characteristic reflections of
SBA-15 disappear entirely. Similar findings have also been reported in the literature for the
synthesis of aminopropyl-functionalized SBA-15 and thiol-functionalized MCM-41 materials via
the co-condensation method, which are attributed to several possible mechanisms [40, 44]. First of
12
all, in acidic condition, some amine groups from organosilane AAPTS protonate to form
zwitterions (−NH3+….−OSi) with silanol groups [45], which prevent the interaction between
surfactant P123 and silicates, thus forming less ordered pore structure. Secondly, herein, the
assembly of SBA-15 mesoporous structure is organized by surfactant P123 in an acidic media via
the intermediate specie (S0H+)(X-I+) [46], in which S0 is surfactant P123, H+ is the hydrogen ion,
X- is the halide anion, and I+ is the protonated Si−OH moiety. In the presence of organosilane
AAPTS, there may be a competition existing between the amine groups and P123, resulting in the
formation of the intermediate (−NH2H+)(X-I+) instead of (S0H+)(X-I+) [47]. Therefore, the addition
of AAPTS in the synthesis solution strongly affects the self-assembly of SBA-15 mesostructure.
On the other side, F- anions, which are added in the synthesis gel, play a catalytic role in
promoting the polymerization of silicates [48]. Because of the small radius and high charge
density of F− anions, the electrostatic interactions among F−, charge-associated P123 (S0H+) and
protonated Si−OH moiety (I+) are much stronger as compared with other anions (e.g. Cl−). This
can enhance the formation of (S0H+)(X-I+), resulting in SBA-15 well-ordered mesostructure [48,
49]. Therefore, in comparison with S15-NN-Fe-0, S15-NN-Fe-0.1 (Figure 1b) can still display
three characteristic diffraction peaks of SBA-15 at low 2. values, despite lower densities. This
indicates that the absorbent synthesized in the gel with a low molar ratio of organosilane in the
presence of F− anions can largely maintain hexagonal mesoporous structure of SBA-15. However,
at a higher AAPTS/TEOS molar ratio, especially above 0.4, d100 spacing of resulting sample is
observed dramatically decreased; and the high order (110) and (200) peaks diminish and finally
disappear. This suggests that it becomes difficult to form well-ordered mesoporous structure of
SBA-15 with an increasing molar ratio of AAPTS, despite with the assistance of F− anions.
13
Figure 2 shows FT-IR spectra of S15-NN-Fe-x (x = 0-0.6) samples. In the FT-IR spectrum of
S15-NN-Fe-0 (Figure 2a), the peak at 950 cm−1 is ascribed to non-condensed silanol Si-OH groups
[50, 51]; and several bands at 1080 cm−1, 800 cm−1, and 465 cm−1 are attributed to the Si-O-Si
asymmetric and symmetric vibrations of condensed silica network. Similar peaks can also be
observed in all of our functionalized silica absorbents, from S15-NN-Fe-0.1 (Figure 2b) to
S15-NN-Fe-0.6 (Figure 2g). With the addition of AAPTS in the synthesis gel, there are several
new peaks appearing in the FT-IR spectra of resulting samples (Figure 2b-g). The bands observed
at 690 cm−1, 1470 cm−1 and 1650 cm−1 are corresponding to –N-H bending vibration and NH2
vibrations [52, 53]. The peaks in the range of 2900-3000 cm−1 are ascribed to alkyl C-H stretching
vibrations. The appearance of those new peaks confirms the successful incorporation of organic
functional groups into the synthesized absorbents via the co-condensation method. The intensities
of these new peaks increase as a greater amount of AAPTS incorporated in the resulting samples,
especially for S15-NN-Fe-0.6 (Figure 2g), which is consistent with the results of N contents
measured in elemental analysis (Table 1).
Figure 3 shows N2 adsorption–desorption isotherms (Figure 3A) of S15-NN-Fe-x (x = 0-0.6)
and their corresponding BJH pore size distribution plots (Figure 3B). Table 1 summarizes BET
surface areas (SBET) and total pore volumes (Vtotal) of S15-NN-Fe-x (x = 0-0.6) samples. The N2
adsorption–desorption isotherm of S15-NN-Fe-0 exhibits type IV isotherm model with H1
hysteresis loop according to IUPAC classification standard (Figure 3Aa), suggesting the sample
with uniform and even mesopores [54]. It can be further seen from the meospore size distribution
of S15-NN-Fe-0 in Figure 3Ba that the peak pore is 5.6 nm; and SBET and Vtotal (Table 1) are
652.53 m2/g and 0.91 cm3/g, respectively. There is a change in the shapes of isotherms observed
14
for the samples prepared with an increase of AAPTS/TEOS molar ratios in Figure 3Ab-g.
According to IUPAC standard, the N2 adsorption–desorption isotherms of S15-NN-Fe-0.1 (Figure
3Ab), S15-NN-Fe-0.2 (Figure 3Ac) and S15-NN-Fe-0.3 (Figure 3Ad) can be classified as type IV
isotherm model with H1 hysteresis loop; whilst the isotherms of S15-NN-Fe-0.4 (Figure 3Ae) and
S15-NN-Fe-0.5 (Figure 3Af) both exhibit type I isotherm model with H2 type hysteresis. The
samples with greater AAPTS show that the position of capillary condensation steps shifts to lower
pressure values, suggesting there is a reduction in the mesopore sizes. This is also confirmed by
the decrease of peak pores as shown in Figure 3Bb-g. In Table 1, SBET and Vtotal are also found to
decrease for the samples synthesized with an elevated AAPTS molar ratio. For instance, SBET and
Vtotal of S15-NN-Fe-0.2 are 382.20 m2/g and 0.66 cm3/g, respectively, as compared with 168.94
m2/g and 0.20 cm3/g for S15-NN-Fe-0.5. Those changes in textural properties can be explained by
several reasons. First of all, after the co-condensation reaction with the presence of organosilane,
organic functional groups may occupy the mesopore channels of resulting materials. In particular,
the presence of an elevated concentration of AAPTS in the synthesis gel perturbs the
self-assembly of mesoporous silica, resulting in the pore shrinkage and less ordered mesoporous
structure. In addition, the impregnation of Fe(III) may cause a reduction of BET surface areas and
pore volumes, due to the occupation of volume within the mesopores of materials. For
S15-NN-Fe-0.6 prepared with the AAPTS/TEOS molar ratio of 0.60, its SBET and Vtotal drop
dramatically to 4.53 m2/g and 0.01 cm3/g (Table 1), which is ascribed to the formation of
disordered pore structure, as established in Figure 1g and further supported by its TEM image
(Figure 5g).
The N contents (mmol/g) and Fe contents (mmol/g) of samples are summarized in Table 1,
15
which are used to evaluate the amounts of incorporated functional groups and positive sites in
S15-NN-Fe-x (x = 0-0.6) samples. The degrees of immobilization are thus derived from the
measured and theoretical N contents of samples, and included in Table 1. Without the addition of
organosilane in the synthesis, S15-NN-Fe-0 does not display any N content, and only 0.0035
mmol/g of Fe3+ can be impregnated into the sample from solution. By increasing AAPTS/TEOS
molar ratios, there is an increase in the N contents, indicating more amino functional groups have
been incorporated into the materials. For instance, the N contents of S15-NN-Fe-0.1,
S15-NN-Fe-0.3, and S15-NN-Fe-0.6 are observed to be 1.32, 2.39 and 4.24 mmol/g, respectively.
However, there is a decrease in the degrees of immobilization accompanied. The degree of
immobilization of S15-NN-Fe-0.1 is 0.47, which is 0.11 and 0.04 greater than those of
S15-NN-Fe-0.3 and S15-NN-Fe-0.6. This is because that only a portion of added diamino silanes
can be incorporated into the resulting absorbents after the co-condensation reaction. Not
surprisingly, for the absorbent prepared with a higher AAPTS/TEOS molar ratio, there is a greater
Fe content observed, suggesting more positive sites are available in the material for phosphate
adsorption. For example, the Fe content of S15-NN-Fe-0.6 is 0.60 mmol/g, which is twofold
higher than that of S15-NN-Fe-0.3.
Figure 4 and 5 shows SEM and TEM images of S15-NN-Fe-x (x = 0-0.60). In Figure 4a-f, all
of samples exhibit a uniform and ropelike shape, which is the typical SBA-15 morphology as
reported in literature [55, 56]. By contrast, in Figure 4g, S15-NN-Fe-0.6 shows an irregular shape.
In Figure 5a–d, the samples exhibit well-ordered mesoporous SBA-15 arrangement of mesopores
and no apparent shrinking of pores are observed. In Figure 5e and f, S15-NN-Fe-0.4 and
S15-NN-Fe-0.5 partially retain the pore ordering, accompanied with a reduced mesoporosity.
16
Whilst S15-NN-Fe-0.6, as shown in Figure 5g, exhibits totally disordered structure without any
mesoporous ordering. All of those findings are consistent with the results in XRD patterns (Figure
1) and nitrogen adsorption-desorption analysis as above-mentioned (Figure 3 and Table 1).
3.2 Phosphate adsorption
Figure 6 shows the degrees of phosphate removal by using the synthesized adsorbents
S15-NN-Fe-x (x = 0-0.6). It clearly shows that the adsorbent S15-NN-Fe-0 without diamino
functional groups has only 1.71% phosphate removal, indicating that it hardly removes phosphate
anions (Figure 6a). This can be explained by its ICP result (Table 1) that only a small amount
(0.0035 mmol/g) of Fe3+ ions can be impregnated into S15-NN-Fe-0, which is a pure SBA-15
material. After organic-functionalization, the resulting Fe-coordinated absorbents possess
significantly enhanced phosphate removals, as compared with S15-NN-Fe-0. Furthermore, by
increasing molar ratios of AAPTS and TEOS from 0.10 to 0.50, the degrees of phosphate removal
of samples increase gradually. For example, the degrees of phosphate removal of S15-NN-Fe-0.1,
S15-NN-Fe-0.3, and S15-NN-Fe-0.5 are 40.19%, 76.05%, and 95.84%, respectively (Figure 6b-f).
It has been shown that by increasing addition of organosilane in the synthesis, our samples are
functionalized with greater loadings of amino functional groups and can partially retain the
mesoporous structures. Due to the greater functionalization degrees, more Fe3+ ions can be
impregnated into the materials to form Fe3+ – en complexes as trapping centers for phosphate. In
particular, we utilize the direct co-condensation method in the synthesis of organic-functionalized
adsorbents, which has been proven as an effective strategy to provide a uniform distribution of
organic groups inside the framepores without the concentration and block of channel entries, as
compared with the post-grafting method [34, 57]. The active sites inside the pores of synthesized
17
absorbents are highly accessible to phosphate anions. Hence, there is an enhancement in degrees
of phosphate removal observed for the samples with increasing AAPTS ratios. Similar results
were also reported in the adsorption of Hg(II) onto the thiol-functionalized MCM-41 which was
fabricated via the co-condensation method [58]. Walcarius and co-workers found that the higher
degree of functionalization led to the formation of poorly ordered materials, whereas the
maximum Hg(II) adsorption could reach. However, when compared to S15-NN-Fe-0.5, the degree
of phosphate removal with the use of S15-NN-Fe-0.6 declines (Figure 6g), although the Fe content
of S15-NN-Fe-0.6 is 0.17 mmol/L greater than that of S15-NN-Fe-0.5 (Table 1). This may be
explained by the formation of totally disordered structure associated with S15-NN-Fe-0.6, in
which active sites available for phosphate anions to reach are limited.
To further evaluate the phosphate adsorption capacity, Figure 7 shows Langmuir (a) and
Freundlich (b) adsorption isotherms of S15-NN-Fe-x (x = 0.1, 0.3 and 0.5). In Figure 7 and Table
2, it shows that both of Langmuir and Freundlich equations can satisfactorily describe the
isotherms experimental data (R2 > 0.9). However, the use of Langmuir equation seems better than
the use of Freundlich equation in describing the adsorption isotherms, because R2 is over 0.994.
This suggests that the observed sorption feature is caused by the monolayer adsorption, which is
similar to the findings reported in the literature by using Al-impregnated mesoporous silicates [59]
and La-doped mesoporous silicates for phosphate removal [60]. As shown in Table 2, quantitative
Langmuir parameters q0 for S15-NN-Fe-0.5, which is 20.7 mg P/g, is significantly greater than
those of S15-NN-Fe-0.1 (4.7 mg P/g) and S15-NN-Fe-0.3 (8.8 mg P/g). This is ascribed to the
higher level of functionalization in S15-NN-Fe-0.5, which can provide more active sites for
phosphate adsorption, when compared with S15-NN-Fe-0.1 and S15-NN-Fe-0.3.
18
Figure 8 shows the adsorption kinetic study with the use of S15-NN-Fe-0.5, which exhibits the
highest degree of phosphate removal among all of our samples in Figure 6, in the solution with
different initial phosphate concentrations, i.e. 0.65, 1.63 and 2.60 mmol/L. In various
concentrations of phosphate solution, S15-NN-Fe-0.5 shows fast adsorption rates and over 75% of
final adsorption capacities reach in the first 1 min. In the following 10 min, the adsorption
capacities increase and almost 90% of final adsorption capacities are achieved. This is ascribed to
the large concentration gradient between bulk solution and adsorbent surface. Moreover, with
partially retained pore structures, S15-NN-Fe-0.5 can allow phosphate anions to access to the
binding sites easily. Between 10 min and 30 min, the adsorption rates become slow and the
intraparticle diffusion mechanism governs these adsorption processes. After 30 min, the phosphate
adsorption processes reach the equilibrium and the adsorption capacities remain constant after 60
min, which may be attributed to the occupancy of sorption sites. The maximum phosphate
adsorption capacities of S15-NN-Fe-0.5 are found to be 16.6, 18.4, and 20.7 mg P/g in 0.65, 1.63
and 2.60 mmol/L of phosphate solution, respectively. As seen, the kinetic study has proven that the
adsorption onto our fabricated absorbent in different concentrations of phosphate solution can
reach equilibrium after 60 min. Therefore, in our batch experiments, the degrees of removal were
recorded after 2 h to ensure the absorbents reached the adsorption equilibrium.
Figure 9 shows the kinetic curves of phosphate adsorption on S15-NN-Fe-0.5 fitted in the
pseudo-second-order model; and the corresponding parameters and correlation coefficients are
listed in Table 3. The fitted pseudo-second-order kinetic models show high correlation coefficients
(R2 = 0.999). An attempt was also conducted to use the pseudo-first-order kinetic model to fit the
kinetic curves, but R2 is much lower, only around 0.650 (Table 3). Thereby, our results suggest
19
that the adsorption on S15-NN-Fe-0.5 in different concentrations of phosphate solution be
chemisorption.
Figure 10 exhibits the degrees of phosphate removal of S15-NN-Fe-0.5 by varying initial pH
values in the range of 2.0 –11.0. The degree of removal is about 68.92% at pH 2.0 and then
increases to 94.67% at pH 3.0. The removal degrees fluctuate at 94.42% from pH 3.0 to 6.0. As
the pH is further increased from 6.0 to 11.0, the degrees of removal decrease sharply from 95.31%
to almost 0%. It is known that phosphate, which is a polyacid, can exist in different ionic species
of H2PO4−, HPO4
2−, and PO43−, depending on the pH of solution [61]. When the pH value is lower
than 2.13, the predominant species of phosphate is the neutral H3PO4, which is weakly attached to
the sites of S15-NN-Fe-0.5. However, Fe3+ may leach out from S15-NN-Fe-0.5 at low pH [28],
followed by the protonation of en ligands. The ammonium moieties are capable of attracting
anionic species by electrostatic forces [62], thus there is 68.92% phosphate removal at pH 2.0.
When pH value is between 2.13 and 7.20, the main species is monovalent H2PO4−. The relatively
high degrees of removal between pH 3.0 and 6.0 indicate that the Fe3+ – en complex centers
provide greater affinity for the single charged phosphate species (H2PO4−). At pH between 7.2 and
11.0, the predominant species of phosphate in aqueous solution is HPO42−. The degrees of removal
for S15-NN-Fe-0.5 dramatically decrease at pH ≥ 7.0, ascribed to the competitive adsorption
between HPO42− and OH−. The other possible reason may be related to the precipitation of Fe3+
when pH is above 7.0. Fe is mainly in its insoluble form Fe(OH)3, which can be hardly trapped by
S15-NN-Fe-0.5.
Figure 11 shows the effects of coexisting anions, including F−, Cl−, NO3−, SO4
2−, and HCO3−,
on the degrees of phosphate removal with the use of S15-NN-Fe-0.5. Without adding any
20
competitive anions, the degree of removal of S15-NN-Fe-0.5 is as high as 92.52%. However, there
is a decrease in the degrees of removal in the presence of competitive anions, which are in the
order: HCO3− >SO4
2− >F− >NO3− >Cl−. This trend of competitive anions on phosphate removals is
in agreement with the results reported by using cationic metal (i.e. Cu, Fe, and La)-coordinated
amino-functionalized MCM-41 [28,30,63]. The decrease of phosphate removals with coexisting
anions may be explained by ion exchange mechanisms, in which the competitive anions may
displace the phosphate ions adsorbed on S15-NN-Fe-0.5 [29]. Furthermore, the affinity of
adsorbent with competing ions determines the orders of the effect of anions on the degree of
removal. Chouyyok et al. suggested that metal-coordinated amino-functionalized mesoporous
silica can bound the anions of higher base strength more strongly and largely than those with
lower base strength [28, 63]. Therefore, among various competitive anions, HCO3− possesses
strongest affinity to the adsorbent and competes most effectively against phosphate adsorption.
Only 1.64% of phosphate anions are removed by S15-NN-Fe-0.5 in the presence of HCO3−. On
the other hand, with the addition of weak base anions Cl− or NO3− , the degrees of removal for
S15-NN-Fe-0.5 are slightly lower, which are 85.71% and 79.44%, respectively. In Figure 11, in
addition to HCO3−, SO4
2- exhibit a relatively strong effect on the phosphate removal, which may
be not only related to its strong base strength, but also the geometry of tetrahedral anions which
would well match the 3-fold symmetry structure of metalated en-functionalized mesoporous silica
[64]. To evaluate its regeneration ability, the desorption kinetic study on the spent adsorbent
S15-NN-Fe-0.5 is performed in 0.010 mol/L NaOH (Figure 12). It is clear that the desorption
process is fast and almost completes within 15 min, which indicates the adsorption for phosphate
anions is reversible and the spent absorbent shows a capacity of regeneration.
21
Table 4 compares the adsorption capacity of S15-NN-Fe-0.5 with other adsorbents used for
phosphate adsorption in literature. In particular, when the pH and temperature are 5 and 35 °C,
S15-NN-Fe-0.5 exhibits higher adsorption capacity than Lanthanum/aluminum pillared
montmorillonite or aluminum pillared montmorillonite [12]. Furthermore, Table 4 demonstrates a
great possibility that our fabricated absorbent could possess a promising performance in the
phosphate removal as compared with other materials. However, some of the adsorption capacities
are not strictly comparable, since the tests are conducted under different experimental conditions,
including pH and temperature [29]. Therefore, our further work at this aspect is under progress.
4. Conclusion
A series of diamino-functionalized mesoporous silica with different functionalization levels
were synthesized via a new F- assisted co-condensation method and the incorporation of Fe3+
cations. Fe-coordinated amino-functionalized absorbents, synthesized in the gel with
AAPTS/TEOS molar ratios from 0.10 to 0.60, possess increasing loadings of amino groups,
accompanied with gradual changes in their textual and structure properties. As compared with the
pure SBA-15 mesoporous silica, all of the amino-functionalized absorbents after the impregnation
of Fe3+ showed enhanced phosphate adsorption capacities. In particular, the sample
S-15-NN-Fe-0.5 possessed a highest adsorption capacity (20.7 mg P/g) in phosphate solution,
whereas S-15-NN-Fe-0 can only remove 1.71 % of phosphate in solution. The phosphate
adsorption equilibrium data were fitted better with the use of Langmuir model than Freundlish
model, indicating the phosphate removal was governed by monolayer adsorption. For
S-15-NN-Fe-0.5, its adsorption kinetic data can well described by the pseudo second-order model
22
with R2 = 0.999. The optimal pH for phosphate adsorption with the use of S15-NN-Fe-0.5 was
between 3.0 and 6.0, and the phosphate absorbed on the spent adsorbent could be desorbed in
0.010 mol/L NaOH.
Acknowledgements
This work was supported by Natural Science Foundation of Guangdong (No.
S2011040001667), the Fundamental Research Funds for the Central Universities (No. 21611310),
the National High Technology Research and Development Program of China (863 Program) (No.
2009AA064401), and the Key Laboratory of Mineralogy and Metallogeny Cooperation
Foundation (No. KLMM20110204).
23
References and Notes
[1]. K.G. Sellner, G.J. Doucette, G.J. Kirkpatrick, J. Ind. Microbiol. Biotechnol. 30 (2003)
383–406.
[2] G.K. Morse, S.W. Brett, J.A. Guy, J.N. Lester, Sci. Total Environ. 212 (1998) 69–81.
[3] G. Zhang, H. Liu, R. Liu, J. Qu, J. Colloid Interface Sci. 335 (2009) 168–174.
[4] P. Ning, H.J. Bart, B. Li, X. Lu, Y. Zhang, J. Environ. Sci. 20 (2008) 670–674.
[5] Z. Hongshao, R. Stanforth, Environ. Sci. Technol. 35 (2001) 4753–4757.
[6] F. Gan, J. Zhou, H. Wang, C. Du, X. Chen, Water Res. 43 (2009) 2907–2915.
[7] H. Li, J. Ru, W Yin, X. Liu, J. Wang, W. Zhang, J. Hazard. Mater. 168 (2009) 326–330.
[8] L. Zeng, X. Li, J. Liu, Water Res. 38 (2004) 1318–1326
[9] J. Das, B.S. Patra, N. Baliarsingh, K.M. Parida, Appl. Clay Sci. 32 (2006) 252–260.
[10] K Karageorgiou, M Paschalis, G N Anastassakis, J. Hazard. Mater. 139 (2007) 447–452.
[11] M.S. Onyango, D. Kuchar, M. Kubota, H. Matsuda, Ind. Eng. Chem. Res. 46 (2007)
894–900.
[12] S. Tian, P. Jiang, P. Ning, Y. Su, Chem. Eng. J.151 (2009) 141–148
[13] W. Huang, S. Wang, Z. Zhu, L. Li, X. Yao, V. Rudolph, F. Haghseresht, J. Hazard. Mater. 158
(2008) 35–42.
[14] A. Ugurlu, B. Salman, Environ. Int. 24 (1998) 911–918.
[15] B. Kostura, H. Kulveitova, J. Lesko, Water Res. 39 (2005) 1795–1802.
[16] X.P. Liao, Y. Ding, B. Wang, B. Shi, Ind. Eng. Chem. Res. 45 (2006) 3896–3901.
[17] B.K. Biswas, K. Inoue, K.N. Ghimire, H. Harada, K. Ohto, H. Kawakita, Bioresour. Technol.
99 (2008) 8685–8690.
24
[18] X. Feng, G.E. Fryxell, L.Q. Wang, A.Y. Kim, J. Liu, K.M. Kemner, Science 276 (1997)
923–926.
[19] M. Puanngam, F. Unob, J. Hazard. Mater. 154 (2008) 578–587.
[20] H. Yoshitake, T. Yokoi, T. Tatsumi, Chem. Mater. 15 (2003) 1713–1721.
[21] Y. Tang, E. Zong, H. Wan, Z. Xu, S. Zheng, D. Zhu. Micropor. Mesopor. Mater. 155 (2012)
192–200.
[22]A.H. Janssen, P. Van Der Voort, A.J. Koster, K.P. De Jong, Chem. Commun. (2002)
1632–1633.
[23] J. Fan, J. Lei, L. Wang, C. Yu, B. Tu, D. Zhao, Chem. Commun. 17 (2003) 2140–2141.
[24] X. Wang, K.S.K. Lin, J.C.C. Chan, S. Cheng, J. Phys. Chem. B 109 (2005) 1763–1769.
[25] Y. Mori, T.J. Pinnavaia, Chem. Mater. 13 (2001) 2173–2178.
[26] L. Beaudet, K.Z. Hossain, L. Mercier, Chem. Mater. 15 (2003) 327–334.
[27] Y. Zhu, H. Li, Q. Zheng, J. Xu, X. Li, Langmuir 28 (2012) 7843–7850
[28]W. Chouyyok, R.J. Wiacek, K. Pattamakomsan, T. Sangvanich, R.M. Grudzien, G.E. Fryxell,
Environ. Sci. Technol. 44 (2010) 3073–3078.
[29]J. Zhang, Z. Shen, W. Shan, Z. Mei, W Wang, J. Hazard. Mater. 186 (2011) 76–83.
[30]J. Zhang, Z. Shen, Z. Mei, S. Li, W. Wang, J. Environ. Sci. 23 (2011) 199–205.
[31]Y. Long, J Yang, B Chen, W. Huang, Y. Zhang, Environ. Sci. Technol. 35 (2012) 15-18 (in
Chinese)
[32] W. Huang, J. Yang, Y. Zhang, Advanced Materials Research. 476-478 (2012) 1969-1973
[33] K. Cassiers, T. Linssen, M. Mathieu, M. Benjelloun, K. Schrijnemakers, P. Van Der Voort, P.
Cool, E.F. Vansant, Chem. Mater. 14 (2002) 2317–2324.
25
[34] T. Yokoi, H. Yoshitake, T. Tatsumi, J. Mater. Chem. 14 (2004) 951–957.
[35] R. Richer, L. Mercier, Chem. Commun. (1998) 1775–1776.
[36] D. Margolese, J.A. Melero, S.C. Christiansen, B.F. Chemelka, G.D. Stucky, Chem. Mater. 12
(2000) 2448–2459.
[37] A.S.M. Chong, X.S. Zhao, J. Phys. Chem. B 107 (2003) 12650–12657.
[38] D.G. Choi, S.M. Yang, J. Colloid Interface Sci. 261 (2003) 127–132.
[39] S. Hao, H. Chang, Q. Xiao, Y. Zhong, W. Zhu, J. Phys. Chem. C 115 (2011) 12873–12882.
[40] S.Y. Hao, Q. Xiao, Y.J. Zhong, W.D. Zhu, H. Yang, Chinese J. Inorg. Chem. 26 (2010)
982–988.
[41] P. Pengthamkeerati, T. Satapanajaru, P. Clularuengoaksorn, Fuel 87 (2008) 2469–2476.
[42] Y.S. Ho, G. McKay, Water Res. 34 (2000) 735–742.
[43] Y.S. Ho, G. Mckay, Process Biochem. 34 (1999) 451–465.
[44] F.O.M. Gaslain, C. DelacÔte, A. Walcarius, B. Lebeau, J. Sol-Gel Sci. Technol. 49 (2009)
112–124.
[45] A. Walcarius, M. Etienne, B. Lebeau, Chem. Mater. 15 (2003) 2161–2173.
[46] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky, Science
279 (1998) 548–552.
[47] X.G. Wang, K.S.K. Lin, J.C.C. Chan, S. Cheng, J. Phys. Chem. B 109 (2005) 1763–1771.
[46] Y.F. Shao, L.Z. Wang, J.L. Zhang, M. Anpo, Micropor. Mesopor. Mater. 86 (2005) 314–322.
[48] D.Y. Zhao, Q.S. Huo, J.L. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc. 120 (1998)
6024–6036.
[50] R.P.W. Scott, Silica Gel & Bonded Phases, Their Production, Properties & Use in LC, Wiley
26
Science, New York, 1993.
[51] L.D. White, C.P.J. Tripp, Colloid Interface Sci. 232 (2000) 400–407.
[52] A.S.M. Chong, X.S. Zhao, J. Phys. Chem. B 107 (2003) 12650–12657.
[53] X.G. Wang, K.S.K. Lin, J.C.C. Chan, S. Cheng, J. Phys. Chem. B 109 (2005), 1763–1769.
[54] M. Kruk, M. Jaroniec, A. Sayari, J. Phys. Chem. B 101 (1997) 583–589.
[55] D.Y. Zhao, J.L. Feng, Q.S. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky,
Science 279 (1998) 548–552.
[56] Y.C. Hsu, Y.T. Hsu, H.Y. Hsu, C.M. Yang, Chem. Mater. 19 (2007) 1120–1126.
[57] L. Mercier, T. J. Pinnavaia, Chem. Mater. 12 (2000) 188–196.
[58] A. Walcarius, C. DelacÔte, Chem. Mater. 15 (2003) 4181-4192
[59] E.W. Shin, J.S. Han, M. Jang, S.H. Min, J.K. Park, R.M. Rowell, Environ. Sci. Technol. 38
(2004) 912–917.
[60] E. Ou, J. Zhou, S. Mao, J. Wang, F. Xia, L. Min, Colloids Surf. A 308 (2007) 47–53.
[61] D.D. Perrin, B. Dempsey, Buffers for pH and metal ions control, Chapman & Hall, London,
1974
[62]S. Hamoudi, R. Saad, K. Belkacemi, Ind. Eng. Chem. Res. 46 (2007) 8806–8812.
[63]S.V. Mattigod, G.E. Fryxell, K.E. Parker, Inorg. Chem. Commum. 10 (2007) 646–648.
[64]G.E. Fryxell, J. Liu, T.A. Hauser, Z.M. Nie, K.F. Ferris, S. Mattigod, M.L. Gong, R.T. Hallen,
Chem. Mater. 11 (1999) 2148–2154
27
Scheme, Table and Figure Captions
Scheme 1. Schematic diagram on the preparation of Fe-coordinated diamino-functionalized silica
absorbents, as well as their phosphate absorption and desorption processes.
Table 1. Chemical compositions and textural characteristics of samples, S15-NN-Fe-0,
S15-NN-Fe-0.1, S15-NN-Fe-0.2, S15-NN-Fe-0.3, S15-NN-Fe-0.4, S15-NN-Fe-0.5, and
S15-NN-Fe-0.6.
Table 2. Langmuir and Freundlich isotherms parameters in the phosphate adsorption by using
S15-NN-Fe-0.1, S15-NN-Fe-0.3, and S15-NN-Fe-0.5.
Table 3. Comparison of the first-order and second-order adsorption rate constants and
experimental values for the sample S15-NN-Fe-0.5 in the phosphate solution with different initial
concentrations (e.g. 0.65, 1.63, and 2.60 mmol/L).
Table 4. Comparison of phosphate adsorption capacities between S15-NN-Fe-0.5 and other
adsorbents reported in literature.
Figure 1. XRD patterns of samples S15-NN-Fe-0 (a), S15-NN-Fe-0.1 (b), S15-NN-Fe-0.2 (c),
S15-NN-Fe-0.3 (d), S15-NN-Fe-0.4 (e), S15-NN-Fe-0.5 (f), and S15-NN-Fe-0.6 (g), respectively.
Figure 2. FT-IR spectra of samples S15-NN-Fe-0 (a), S15-NN-Fe-0.1 (b), S15-NN-Fe-0.2 (c),
S15-NN-Fe-0.3 (d), S15-NN-Fe-0.4 (e), S15-NN-Fe-0.5 (f), and S15-NN-Fe-0.6 (g),respectively.
Figure 3. (A) N2 adsorption-desorption isotherms of synthesized samples, S15-NN-Fe-0 (a),
S15-NN-Fe-0.1 (b), S15-NN-Fe-0.2 (c), S15-NN-Fe-0.3 (d), S15-NN-Fe-0.4 (e), S15-NN-Fe-0.5
(f) and S15-NN-Fe-0.6 (g); and (B) their corresponding BJH pore size distribution plots.
Figure 4. SEM images of samples, S15-NN-Fe-0 (a), S15-NN-Fe-0.1 (b), S15-NN-Fe-0.2 (c),
S15-NN-Fe-0.3 (d), S15-NN-Fe-0.4 (e), S15-NN-Fe-0.5 (f), and S15-NN-Fe-0.6 (g),respectively.
Figure 5. TEM images of samples, S15-NN-Fe-0 (a), S15-NN-Fe-0.1 (b), S15-NN-Fe-0.2 (c),
28
S15-NN-Fe-0.3 (d), S15-NN-Fe-0.4 (e), S15-NN-Fe-0.5 (f), and S15-NN-Fe-0.6 (g),respectively.
Figure 6. The degrees of phosphate removal of S15-NN-Fe-0 (a), S15-NN-Fe-0.1 (b),
S15-NN-Fe-0.2 (c), S15-NN-Fe-0.3 (d), S15-NN-Fe-0.4 (e), S15-NN-Fe-0.5 (f), and
S15-NN-Fe-0.6 (g).
Figure 7. (a) Langmuir adsorption isotherms and (b) Freundlich adsorption isotherms of
S15-NN-Fe-0.1, S15-NN-Fe-0.3, and S15-NN-Fe-0.5.
Figure 8. Effect of contact time and initial concentrations (e.g. 0.65, 1.63, and 2.60 mmol/L) of
phosphate solution on the adsorption capacities of S15-NN-Fe-0.5.
Figure 9. Pseudo-second-order plots for the phosphate adsorption onto S15-NN-Fe-0.5.
Figure 10. Effect of pH on the degrees of removal of S15-NN-Fe-0.5 in 0.065 mmol/L solution.
Figure 11. Effect of co-existing anions on the degrees of phosphate removal of S15-NN-Fe-0.5.
Figure 12. Desorption kinetics of phosphate for the spent S15-NN-Fe-0.5 in 0.010 mol/L NaOH
solution.
29
Scheme 1.
Table 1
Samples N content
(mmol/g)
Degree of
immobilization of N
Fe content
(mmol/g)
SBET
(m2/g)
VTotal
(cm3/g)
S15-NN-Fe-0 0 0 0.0035 652.53 0.91
S15-NN-Fe-0.1 1.32 0.47 0.13 424.14 0.79
S15-NN-Fe-0.2 1.88 0.39 0.25 382.20 0.66
S15-NN-Fe-0.3 2.39 0.36 0.30 368.04 0.62
S15-NN-Fe-0.4 2.43 0.31 0.31 351.96 0.43
S15-NN-Fe-0.5 2.73 0.30 0.43 168.94 0.20
S15-NN-Fe-0.6 4.24 0.43 0.60 4.53 0.01
Table 2
Langmuir Freundlich Samples
q0 (mg/g) KL (L/mg) R2 n KF(mg/g) R2
S15-NN-Fe-0.1 4.7 0.241 0.994 3.1 1.3 0.928
S15-NN-Fe-0.3 8.8 0.420 0.995 4.8 3.7 0.964
S15-NN-Fe-0.5 20.7 0.467 0.994 3.5 7.1 0.929
Table 3
First-order kinetics Second-order kinetics Initial
concentration
C0 (mmol/L)
qe (exp)
(mg P/g) k1
(min−1)
qe(cal)
(mg/g) R2
k2
(g/(mg min))
qe (cal)
(mg/g) R2
0.65 16.6 0.0191 3.8 0.675 0.0413 16.7 0.999
1.63 18.4 0.0183 4.4 0.661 0.0251 18.4 0.999
2.64 20.7 0.0244 4.1 0.612 0.0357 20.7 0.999
30
Table 4
Adsorbents pH Temperature (°C) Adsorption capacity
(mg P/g) References
Fe oxide tailing 6.7-6.8 20 12.7 [8]
Lanthanum/aluminum pillared
montmorillonite 5.0 35 9.8 [12]
Aluminum pillared montmorillonite 5.0 35 7.9 [12]
Red mud 5.5 40 0.6 [13]
ZiO2 functionalized SBA-15 6.2 25 14.7 [21]
Fe-coordinated amino-functionalized
MCM-41 5.0 room temperature 14.3 [28]
La-coordinated amino-functionalized
MCM-41 7.0 35 16.9 [29]
Fe-coordinated amino-functionalized
MCM-41 7.0 35 16.6 [30]
S15-NN-Fe-0.5 5.0 35 20.7 Present work
Figure 1.
31
Figure 2.
Figure 3.
32
Figure 4.
33
Figure 5.
Figure 6.
34
Figure7.
Figure 8.
Figure 9.
35
Figure 10.
Figure 11.
Figure 12.
36
37
� Adsorbents with different functional levels were synthesized via a simple one-pot method.
� The effect of functionalization levels were investigated in detail.
� Both the framework structure and functionalization levels can affect their adsorption
performance.
� The adsorbents with optimum functional level show effective adsorption capacities.